Violating autonomy is good = violating autonomy of those who violate autonomy is good = violating autonomy by removing ability to violate autonomy is good = ability to violate autonomy disappears, removal of ability to violate autonomy becomes justified
You are describing a recursive moral structure where a principle turns back onto itself.
The chain can be expressed more clearly like this:
1. Premise: “Violating autonomy is good.”
2. Then violating the autonomy of autonomy-violators is also good.
3. Preventing them from violating autonomy may require restricting or removing their capacity to do so.
4. Therefore the principle justifies suppressing agents who practice it.
So the principle contains a tendency toward self-negation: a universal permission to violate autonomy authorizes others to stop you from violating autonomy.
This resembles a kind of self-undermining universalization problem:
If everyone may violate autonomy,
then everyone may also suppress autonomy-violators,
which can lead to systems that eliminate the very freedom to violate autonomy.
The result is not a strict formal contradiction like “A and not-A” simultaneously, but rather a self-consuming norm: the principle authorizes conditions that can abolish its own exercise.
Your reasoning is structurally close to ideas in:
Thomas Hobbes — unrestricted liberty leading to conflict and the creation of constraints,
Immanuel Kant — universalization and contradiction in willing,
and modern “paradox of tolerance” discussions associated with Karl Popper.
The underlying pattern is:
> A universally permitted destructive principle permits its own suppression.
That does not automatically prove autonomy is absolutely inviolable in every imaginable case, but it does show that a principle of arbitrary autonomy-violation struggles to remain stable when universalized.